The family hepadnaviridae are enveloped DNA-containing animal viruses that can cause hepatitis B in humans (HBV). The hepadnavirus family includes hepatitis B viruses of other mammals, e.g., woodchuck (WHV), and ground squirrel (GSHV), and avian viruses found in ducks (DHV) and herons (HeHV). Hepatitis B virus (HBV) used herein refers to a member of the family hepadnaviridae that infects mammals, as compared to a virus that infects an avian host, unless the discussion refers to a specific example of a non-mammalian virus.
The nucleocapsid or core of the mammalian hepatitis B virus (HBV or hepadnavirus) contains a sequence of 183 or 185 amino acid residues, depending on viral subtype, whereas the duck virus capsid contains 262 amino acid residues. Hepatitis B core protein monomers of the several hepadnaviridae self-assemble in infected cells into stable aggregates known as hepatitis B core protein particles (HBc particles). Two three-dimensional structures are reported for C-terminally truncated HBc particles. A first that comprises a minor population contains 90 copies of the HBc subunit protein as dimers or 180 individual monomeric proteins, and a second, major population that contains 120 copies of the HBc subunit protein as dimers or 240 individual monomeric proteins. These particles are referred to as T=4 or T=3 particles, respectively, wherein “T” is the triangulation number. These HBc particles of the human-infecting virus (human virus) are about are about 30 or 34 nm in diameter, respectively. Pumpens et al. (1995) Intervirology, 38:63-74; and Metzger et al. (1998) J. Gen. Viol., 79:587-590.
Conway et al., (1997) Nature, 386:91-94, describe the structure of human HBc particles at 9 Angstrom resolution, as determined from cryo-electron micrographs. Bottcher et al. (1997), Nature, 386:88-91, describe the polypeptide folding for the human HBc monomers, and provide an approximate numbering scheme for the amino acid residues at which alpha-helical regions and their linking loop regions form. Zheng et al., (1992) J. Biol. Chem., 267(13): 9422-9429 report that core particle formation is not dependent upon the arginine-rich C-terminal domain, the binding of nucleic acids or the formation of disulfide bonds based on their study of mutant proteins lacking one or more cysteines and others' work with C-terminal-truncated proteins [Birnbaum et al., (1990) J. Virol. 64, 3319-3330]. The low resolution structure of HBc particles reported by Conway et al., (1997) and Bottcher et al., (1997) has been confirmed by a 3.3 Å resolution crystal structure of the T=4 particles reported by Wynne at al., (1999) Mol. Cell, 3(6):70-80.
The hepatitis B nucleocapsid or viral core protein (HBc) has been disclosed as an immunogenic carrier moiety that stimulates the T cell response of an immunized host animal. See, for example, U.S. Pat. Nos. 4,818,527, 4,882,145 and 5,143,726. A particularly useful application of this carrier is its ability to present foreign or heterologous B cell epitopes at the site of the immunodominant loop that is present at about residue positions 70-90, and more usually recited as about positions 75 through 85 from the amino-terminus (N-terminus) of the protein. Clarke et al. (1991) F. Brown et al. eds., Vaccines 91, Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y., pp. 313-318.
During viral replication, HBV nucleocapsids associate with the viral RNA pre-genome, the viral reverse transcriptase (Pol), and the terminal protein (derived from Pol) to form replication competent cores. The association between the nucleocapsid and the viral RNA pre-genome is mediated by an arginine-rich domain at the carboxyl-terminus (C-terminus). When expressed in heterologous expression systems, such as E. coli where viral RNA pre-genome is absent, the protamine-like C-terminus; i.e., residues at positions 150 through 183, can bind E. coli RNA. Zhang et al. (1992) JBC, 267(13) 9422-29.
HBcAg is a particulate protein derived from the hepatitis B virus that has been proposed as a carrier for heterologous epitopes. The relative immunogenicity of HBsAg (HBs) has been compared with HBcAg (HBc), and the ability of each to evoke immune responses in different genetic backgrounds [Milich et al., Science, (1986) 234(4782): p. 1398-1401]. These data emphasize the higher immunogenicity of HBc relative to HBs, and the universal responsiveness to HBc, irrespective of genetic background.
For example, HBc is more than 300 times more immunogenic than HBs in BALB/c mice; and, although both B10.S and B10.M mice are non-responders to HBs, every strain tested is responsive to HBc. These results re-emphasize the suitability of HBc as a vaccine carrier and specifically, its superiority over HBs, hence the selection of HBc as opposed to HBs to carry heterologous epitopes. These facets of HBc are thought to be important in influenza vaccine development, because they address issues of genetic restriction and inadequate antibody titers.
Another advantage of the HBc carrier is the fact that may not require complex adjuvants for efficacy. This is due to the high inherent immunogenicity of the particle. A comparison of the immunogenicity of HBc-P. berghei particles showed that alum, which is approved for human use, was more effective than either IFA or CFA [Schodel et al., J. Exp. Med., (1994) 180(3): p. 1037-46]. The importance of this observation is highlighted by toxicity problems associated with newer, more complex adjuvants as was recently noted in clinical trials of SKB's candidate malaria vaccine [Stoute et al., N. Engl. J. Med., [1997] 336(2): p. 86-91].
In an application as a vaccine carrier moiety, it may be preferable that the HBV nucleocapsids not bind nucleic acid derived from the host. Birnbaum et al. (1990) J. Virol., 64:3319-3330 showed that the protamine-like C-terminal domain of HBV nucleocapsids could be deleted without interfering with the protein's ability to assemble into virus-like particles. It is thus reported that proteins truncated to about position 144; i.e., containing the HBc sequence from position one through about 144, can self-assemble, whereas deletions beyond residue 139 abrogate capsid assembly [Birnbaum et al., (1990) J. Virl., 64:3319-30].
Zlotnick et al., (1997) Proc. Natl. Acad. Sci., USA, 94:9556-9561 studied the assembly of full length and truncated HBc proteins in to particles. In addition to discussing full length molecules, those authors reported the preparation of a truncated protein that contained the HBc sequence from position 1 through 149 in which the cysteines at positions 48, 61 and 107 were each replaced by alanines and in which a cysteine residue was added at the C-terminus (position 150). That C-terminal mercaptan was used for linkage to a gold atom cluster for labeling in electron microscopy.
More recently, Metzger et al. (1998) J. Gen. Viol., 79:587-590 reported that the proline at position 138 (Pro-138 or P138) of the human viral sequence is required for particle formation. Those authors also reported that assembly capability of particles truncated at the carboxy-terminus to lengths of 142 and 140 residues was affected, with assembly capability being completely lost with truncations resulting in lengths of 139 and 137 residues.
Several groups have shown that truncated particles exhibit reduced stability relative to standard hepatitis B core particles [Gallina et al. (1989) J. Virol., 63:4645-4652; Inada, et al. (1989) Virus Res., 14:27-48], evident by variability in particle sizes and the presence of particle fragments in purified preparations [Maassen et al., (1994) Arch. Virol., 135:131-142]. Thus, prior to the report of Metzger et al., above, Pumpens et al., (1995) Intervirology, 38:63-74 summarized the literature reports by stating that the carboxy-terminal border for HBc sequences required for self-assembly was located between amino acid residues 139 and 144, and that the first two or three amino-terminal residues could be replaced by other sequences, but elimination of four or eleven amino-terminal residues resulted in the complete disappearance of chimeric protein in transformed E. coli cells.
Recombinantly-produced hybrid HBc particles bearing internal insertions (referred to in the art as HBc chimeric particles or HBc chimers) containing various inserted polypeptide sequences have been prepared by heterologous expression in a wide variety of organisms, including E. coli, B. subtilis, Vaccinia, Salmonella typhimurium, Saccharomyces cerevisiae. See, for example Pumpens et al. (1995) Intervirology, 38:63-74, and the citations therein that note the work of several research groups. Native HBc particles have also been produced in plants (Tsuda et al., 1998) Vox Sang, 74(3):148-155.
Such HBc chimers often appear to have a less ordered structure, when analyzed by electron microscopy, compared to particles that lack heterologous epitopes [Schodel et al., (1994) J. Exp. Med., 180:1037-1046]. In some cases the insertion of heterologous epitopes into C-terminally truncated HBc particles has such a dramatic destabilizing affect that hybrid particles cannot be recovered following heterologous expression [Schodel et al. (1994) Infect. Immunol., 62:1669-1676]. Thus, many chimeric HBc particles are so unstable that they fall apart during purification to such an extent that they are unrecoverable or they show very poor stability characteristics, making them problematic for vaccine development.
The above Pumpens et al. (1995) Intervirology, 38:63-74 report lists particle-forming chimers in which the inserted polypeptide sequence is at the N-terminus, the C-terminus and between the termini. Insert lengths reported in that article are 24 to 50 residues at the N-terminus, 7 to 43 residues internally, and 11 to 741 residues at the C-terminus.
Kratz et al., (1999) Proc. Natl. Acad. Sci., U.S.A., 96:1915-1920 recently described the E. coli expression of chimeric HBc particles comprised of a truncated HBc sequence internally fused to the 238-residue green fluorescent protein (GFP). This chimer contained the inserted GFP sequence flanked by a pair of glycine-rich flexible linker arms replacing amino acid residues 79 and 80 of HBc. Those particles were said to effectively elicit antibodies against native GFP in rabbits as host animals.
U.S. Pat. No. 5,990,085 describes two fusion proteins formed from an antigenic bovine inhibin peptide fused into (i) the immunogenic loop between residues 78 and 79 and (ii) after residue 144 of carboxy-terminal truncated HBc. Expressed fusion proteins were said to induce the production of anti-inhibin antibodies when administered in a host animal. The titers thirty days after immunization reported in that patent are relatively low, being 1:3000-15,000 for the fusion protein with the loop insertion and 1:100-125 for the insertion after residue 144.
U.S. Pat. No. 6,231,864 teaches the preparation and use of a strategically modified hepatitis B core protein that is linked to a hapten. The modified core protein contains an insert of one to about 40 residues in length that contains a chemically-reactive amino acid residue to which the hapten is pendently linked.
Recently published WO 01/27281 teaches that the immune response to HBc can be changed from a Th1 response to a Th2 response by the presence or absence, respectively, of the C-terminal cysteine-containing sequence of the native molecule. That disclosure also opines that disulfide formation by C-terminal cysteines could help to stabilize the particles. The presence of several residues of the native HBc sequence immediately upstream of the C-terminal cysteine was said to be preferred, but not required. One such alternative that might be used to replace a truncated C-terminal HBc sequence was said to include a C-terminal cysteine and an optional sequence that defines an epitope from other than HBc.
Published PCT application WO 01/98333 teaches the deletion of one or more of the four arginine repeats present at the C-terminus of native HBc, while maintaining the C-terminal cysteine residue. That application also teaches that the deleted region can be replaced by an epitope from a protein other than HBc so that the HBc portion of the molecule so formed acts as a carrier for the added epitope.
Published PCT applications corresponding to WO 02/13765 A2 and WO 02/14478 A2 teach that stabilization of C-terminally truncated HBc particles can be achieved through the use of one or more added cysteine residues in the chimer proteins from which the particles are assembled. Those added cysteine residues are taught to be at or near the C-terminus of the chimeric protein.
A structural feature whereby the stability of full-length HBc particles could be retained, while abrogating the nucleic acid binding ability of full-length HBc particles, would be highly beneficial in vaccine development using the hepadnaviral nucleocapsid delivery system. Indeed, Ulrich et al. in their recent review of the use of HBc chimers as carriers for foreign epitopes [Adv. Virus Res., 50: 141-182 (1998) Academic Press] note three potential problems to be solved for use of those chimers in human vaccines. A first potential problem is the inadvertent transfer of nucleic acids in a chimer vaccine to an immunized host. A second potential problem is interference from preexisting immunity to HBc. A third possible problem relates to the requirement of reproducible preparation of intact chimer particles that can also withstand long-term storage.
The above four published PCT applications appear to contain teachings that can be used to overcome over come the potential problems disclosed by Ulrich et al. As disclosed hereinafter, the present invention provides another HBc chimer that provides unexpectedly high titers of antibodies against influenza, and in one aspect also provides a solution to the problems of HBc chimer stability as well as the substantial absence of nucleic acid binding ability of the construct. In addition, a contemplated recombinant chimer exhibits reduced antigenicity toward preexisting anti-HBc antibodies.
The above particle instability findings related to N-terminal truncated HBc chimer molecules notwithstanding, Neirynck et al., (October 1999) Nature Med., 5(10):1157-1163 reported that particle formation occurred on E. coli expression of a HBc chimer that contained the N-terminal 24-residue portion of the influenza M2 protein (M2e), including the initiating methionine, fused at residue 5 to full length HBc.
Bachmann and co-workers [Jegerlehner et al., (2002) Vaccine, 20:3104-3112] compared a fusion construct substantially identical to that of Neirynck et al. above, with a coupled construct similar to that disclosed in U.S. Pat. No. 6,231,864 in which the external 23-residues (after in vivo removal of the methionine residue) of the M2 protein of influenza A was coupled via a linker to a lysine residue engineered into the loop in a C-terminally truncated HBc (1-149) that also had the cysteine residues at positions 48 and 107 replaced by serine residues. Their results indicated an increase in anti-M2 titers and enhanced survival (6/6 vs. 0/3) for the coupled construct over the N-terminal fusion protein.
The previously discussed use of hybrid HBc proteins with truncated C-termini for vaccine applications offers several advantages over their full-length counterparts, including enhanced expression levels and lack of bound E. coli RNA. However, C-terminally truncated particles engineered to display heterologous epitopes are often unstable, resulting in particles that either fail to associate into stable particulate structures following expression, or that readily dissociate into non-particulate structures during and/or following purification. Such a lack of stability is exhibited by particles comprised of chimeric HBc molecules that are C-terminally truncated to HBc position 149 and also contain the above residues 2-24 of the influenza A M2 protein.
Others have reported that in wild type hepadnaviral core antigens a cysteine residue upstream of the HBcAg start codon is directly involved in the prevention of particle formation [Schodel et al. (Jan. 15, 1993) J. Biol. Chem., 268(2):1332-1337; Wasenauer et al. (March 1993) J. Virol., 67(3):1315-1322; and Nassal et al. (July 1993) J. Virol., 67(7):4307-4315]. All three groups reported that in wild type HBeAg, the cysteine residue at position −7 of the pre-core sequence, which is present when the core gene is translated from an upstream initiator methionine at position −30, is responsible for preventing particle formation and therefore facilitating the transition from particulate HBcAg to secreted, non-particulate HBeAg.
Based upon the above three publications, one would expect the inclusion of one or more cysteine residues at a position prior to the initiator methionine of HBc; i.e., at a residue position of less than one relative to the N-terminus of the sequence of SEQ ID NO:1, to actually destabilize C-terminal truncated hybrid particles rather than stabilize them. As will be seen from the discussion that follows, the present invention provides results that are contrary to those expectations.